Design and Improvement of e-Collab Classroom as Learning Support System on Intelligent System Subject Electrical Department, Universitas Negeri Malang

Practicum activity on Intelligent System requires a computer laboratory with have a Graphics Processing Unit (GPU) computer spesification to increase the speed of computing. A computer with a GPU is a high-tech and expensive device. The amount of practicum equipment needed adjusts the number of students. Especially in practicums held in parallel, the users will increase several times. In addition, this device requires assistance with training in operating procedures, and requiring longer study time. One way to overcome this problem is a cloud computing device as a practicum media. The students and lecturers interact online with practicum media, online compilers, and GPU server facilities for computing. Online classes that are collaborated with learning media are called e-Collab Classroom. The student ability of computer programming is very diverse so that it has the potential to obstruct practical activities. This is used to minimize this condition and unite the platforms used in each topic contained in the Smart Systems Course. This research improve e-Collab Classroom which can be accessed from computer browser to be easily access by students from one gate mobile application. Designing e-Collab based on mobile applications for Intelligent system subjects with the topic of practical activities, namely Convolutional Neural Network (CNN). The topic of CNN was chosen as a specific topic that represented the entire simulation of practical learning activities with cloud computing. The development of e-Collab is carried out using a prototyping model consisting of the identification of needs and objects, the identification of risks, prototype design, the experimental stage, the testing phase, and the evaluation stage. The results of the development research are the CNN jobsheet, online classroom system, and mobile application.